Building my new BMS.

[fechter]

When you unplug the charger and break the jumper connection the main LED goes out and the circuit drains about 2 uA. With the jumper connected, the control circuit draw around 20mA (but only from the first 5 cells).

You don't want the control circuit to be on when there is no charger.

The rest of the circuit drains a very low amount. I think it would be OK for several months (if not longer). If you want to store for long periods, simply unplug the BMS circuit from the batteries.

 

[Patriot]

When I unplug the charger, and disconnect the 3-pin connector, the main LED stays ON.

There is no connection between the two charging GND wires or on the board, so why is the main LED still ON?

If I want to disconnect the POS side to deenergize the whole board, I need to disconnect the 17-pin connector, and the POS lead to the battery pack. I don't want to have to disconnect the 17-pin connector all the time. HOwever, I don't want to do this. I want to keep it all connected for use, so I have LVC capable.

So, how do I deenergize this thing to make the main LED go off?

 

[Icewrench]

The small ground wire needs to be disconnected from all other grounds.

On mine sooner or probably later I will use an on / off switch on that small lead.

 

[Patriot]

The small ground wire needs to be disconnected from all other grounds.

On mine sooner or probably later I will use an on / off switch on that small lead.

fecter, GGoodrum.... HELP!!!!

What I want to know, is do I NEED to switch it off like that all the time?

fecter said that the main LED should go off when I disconnect the charger. So, why isn't my main RED LED going off?

I want it to use the LVC feature when riding. How can that work if I disconnect the main FET GND lead to the pack?

Does the small GND lead for Channel-1 going through my 17-pin connector do the job for the LVC feature?

If I do install a small bullet connector to the main FET GND lead, it won't harm the circuit with all the POS leads attached, will it?

 

[Icewrench]

This is my small wire set up. My charger is always connected to the even newer bms.
Not sure why yours is on, mine is red with small wires connected and
charger off or charging and goes green at full charge and stays green until the small wire connection is opened.
Does this help? If not Never mind.

 

[fechter]

I then noticed that the third connection (THE EXTRA GROUND) was not connected inside the 3-pin connector. It was left blank.

So, I wired in a jumper.

I then plugged it in, and the battery immediately jumped to 54.2vdc and climbing. The fan

Put the jumper on the charger side of the plug so it automatically disconnects when you unplug the charger.

 

[Patriot]

I did put it on the charger side.

Do I need to install a switch on the FET GND wire too?

That's what ejonesss said in GGoodrum's BMS thread.

However, the schematic drawing doesn't show a disconnect or switch of some sort is needed.

 

[fechter]

If you disconnect the small ground wire, the control LED should go out. If it stays on, then I would suspect the switch transistor fried or has a short. View attachment bms01.jpg

 

[Patriot]

Yes, the ground wires are seperate, and the LED stays on. HOwever, the BMS seems to be working properly, as all cells are balancing.

ejonesss said I need to install a switch on the GND lead going to the battery pack.

Is this true?

Will the LVC still work with the if I do this?


Oh wait, if I do this, then the small GND lead for Channel 1 will still be connected, and the LED will still stay lit.

Oh man, something is wrong then.

The switch transistor you say?

Which one is that?

How can I check to see if it is faulty?

 

[fechter]

If the control LED stays on when the wires are separated, then something is wrong.

I don't have any schematics on this computer, but if you follow the skinny charger (-) wire, it goes to a resistor that goes to the transistor in question. I guess it's the one closest to the FET. If you're lucky, the transistor is just inserted backward or something easy to fix.

Check the traces around it for bridges. If clear, then I think you need to desolder it and verify the control LED does not come on when you attach the batteries.

 

[Patriot]

Does it look like it's installed correctly?

According to the little red/white component drawing, the position looks right.

I don't get it. Why would it get fried?

I don't think I did anything to harm it.

Also, it has taken several hours for the last few cells to balance and light up the LED's. Those were almost fully discharged when I got them in the mail. I'm glad I finally charged them.

 

[Patriot]

I removed the 3906 transistor on the top of the pic. The one next to the FET. When I hooked the battery back up, the LED stayed off. I then touched it to the board, and the LED came back on as before. I then flipped it over (180 deg), and touched it to the board. The LED lit up again, but faintly, about half the brightness.

I guess it must be a bad transistor then?


Now I have a bigger problem.

Being a fool, I wondered if the LED would come on when the GND wires are jumpered to energize the circuit. When I plugged in the charger connector (but not plugged into wall outlet), it made a tiny POP, and I saw a little smoke in the area of where the transistor was removed. Something else next to it smoked, but I don't know what. It's becoming obvious I have no idea what I'm doing.

I immediately unplugged the connector, but some damage has been done, and I have no idea what it is.

I then touched the transistor back in place like before, and the LED lit up again, just as before. So, that part is doing the same. But, now I'm afraid to install a new transistor, because I POPPED/SMOKED something.

I should never have taken on this project. I have very limitted knowledge on how to build/troubleshoot/repair circuits. I fear I may have just wasted $120.00, and no ebike to ride.

A cheesey emoticon does not exist for what I feel right now. If there was an emotional state that describes a nuclear meltdown, that would be about it.

 

[fechter]

You are SO close to having this thing working. Don't give up.


There should be nothing going on when you touch the ground wires together and the transistor is removed. The fact that something snapped when you did indicates some kind of wiring malfunction. The parts look properly installed in the picture. Make sure your transistor is really a 3906

Again, hold the board up to a light and check the bottom side carefully for unwanted bridges around the transistor, diode and resistor.

Check the diode D2 with an ohm meter in both directions (diode check function would be better). It should have a high resistance in one direction and a low resistance in the other direction. Bad ones usually look shorted. You might just replace it anyway.

Depending on what kind of wiring malfunction is going on there, you could possibly blow out U1, but I doubt it. You can measure the output of it.

 

[Patriot]

Yes, it is a 3906.

I also checked all the pathways on the back and front again. I just can't find anything out of the ordinary.


Now, concerning the D2 diode.

I checked resistance. One direction is infinite using both digital and analog meters.

The other direction has me perplexed.

Using my digital meter, I got 161kohm. Weird.

So, I tried my old analog meter. It reads 2.5ohms. Getting close to 0.

I think my digital isn't reaing it right, because the analog seems more reliable for these parts for some reason.

Also, with my digital meter, I'm getting 350kohm between the two charger GND wires. Shouldn't these be infinite without the jumper?

-----------------------------------------------------------------------------------------------------------------
Ok, I just back from Radio Shack. I bought a whole pack of 2N3906-PNP transistors.

However, they all have differant numbers on the back.

2N3906-338

2N2907 A331

2N4403 052

Are these acceptable alternates?

They all have straight leads, so I'd have to spread them a little to fit in the board.


Should I plug it in, and touch one of the new transistors to see what happens?

 

[fechter]

Any of those transistors should be fine.

I'd try just touching the leads to the holes and see what happens. Solder is more work to undo.

The diode measurements sound OK. Unless you have a diode check function on your meter, the forward measurement always looks goofy.

 

[Patriot]

Any of those transistors should be fine.

I'd try just touching the leads to the holes and see what happens. Solder is more work to undo.

The diode measurements sound OK. Unless you have a diode check function on your meter, the forward measurement always looks goofy.

I've been using my analog meter, and I seem to get better readings for some reason. I checked D2, and 2.5ohms one direction, 5ohms the other direction. Keep in mind the other side goes backward through the gatedriver/capacitors, etc. So I think it's ok as well. All of the channel diodes read similar, so I think they are ok.

I'll hook it up, and see what happens.

 

[Patriot]

The same is happening with the new transistor as with the old one. When I install it correctly, the red LED comes on, and when I flip it over, the LED comes on faintly. As far as I know, the LED shouldn't come on at all.

Something is wrong. And now I don't think it was the transistor.

I still can find any bridges pathways or faults in the soldering.

 

[bearing]

Bipolar transistors works backwards, but not very good, thats why the LED lights up faintly.

 

[Patriot]

Bipolar transistors works backwards, but not very good, thats why the LED lights up faintly.

But I was told that when installed the right way, it should not light up at all.

What the heck is going on?

 

[fechter]

Don't put the transistor in backward on purpose, it could possibly be damaged.

Try this:

Put the transistor in the right direction and leave the small ground wire disconnected.
Power up and measure the voltage from the small ground wire to the large one.

Normally, this should be around 14v with the small ground disconnected.

 

[fechter]

See if you can get a nice close up picture of the bottom of the board around where the that transistor goes and post it.

 

[Patriot]

^^^Ok.

I don't have to desolder it from the board do I? It just goes to the 3-pin plug, and is not connected to anything.

Can't I just use my meter and touch the two GND points to see if there is voltage?

 

[Patriot]

Ok, here they are. I've traced everything to the best of my ability on the schematic, and I can't find any mistakes. Unless you can see something I missed. Hence, here's the pics.

 

[Patriot]

Ok,

I lightly soldered in the new transistor for testing.

I powered up. The LED came on. Bright red.

I then checked for voltage between the two charger GND points, and got 12.4vdc.

Also, I noticed after testing, the main LED is now getting dimmer and flickering.

It's as if it's trying to go out, but keeps staying on for some reason.

2mins later: Now, it is faint, but steady.

When I check voltage between the charger GND points, the LED gets bright again, then dim when I remove the meter probes.

Here's a pic.

I wonder if I should instal another transistor. This is the one I flipped over during testing. Maybe I damaged it.

 

[Patriot]

I installed another NEW transistor, and powered up. The other testing transistor must have been messed up, because....


THE LED is OFF!!!!!

I assume this is good.